Cognitive

Understand how AI works.

A visual reference for university students and practitioners, from the mathematics of learning to language models, agents and AI security.

Let's go

Contents

Foundations

  1. Defining Intelligence
  2. AI, Machine Learning, and Deep Learning
  3. Data Representation
  4. Probability and Uncertainty
  5. Machine Learning Paradigms

Neural Networks

  1. The Artificial Neuron
  2. Gradient Descent
  3. Backpropagation
  4. Generalization and Overfitting
  5. Convolutional Networks and Vision
  6. Recurrent Networks and Sequences
  7. Reinforcement Learning

Language Models

  1. Tokenization
  2. Embeddings
  3. The Transformer
  4. Large Language Models
  5. Instruction Tuning and RLHF
  6. Sampling and Decoding
  7. Interpretability

Generative and Multimodal Models

  1. Diffusion Models
  2. Multimodal Models
  3. Reasoning Models
  4. Model Compression and Edge AI

Applied AI Systems

  1. Retrieval-Augmented Generation
  2. AI Agents and Tool Use
  3. Model Context Protocol
  4. Multi-Agent Systems
  5. Hallucination and Evaluation

LLM Application Security

  1. AI Threat Modeling
  2. LLM01: Prompt Injection
  3. LLM02: Sensitive Information Disclosure
  4. LLM03: Excessive Agency
  5. LLM04: Supply Chain
  6. LLM05: Data and Model Poisoning
  7. LLM06: Unbounded Consumption
  8. LLM07: Misinformation
  9. LLM08: Hidden Context Exposure
  10. LLM09: Vector and Embedding Weaknesses
  11. LLM10: Improper Output Handling

Agentic AI Security

  1. ASI01: Agent Goal Hijack
  2. ASI02: Tool Misuse and Exploitation
  3. ASI03: Identity and Privilege Abuse
  4. ASI04: Agentic Supply Chain Vulnerabilities
  5. ASI05: Unexpected Code Execution
  6. ASI06: Memory and Context Poisoning
  7. ASI07: Insecure Inter-Agent Communication
  8. ASI08: Cascading Failures
  9. ASI09: Human-Agent Trust Exploitation
  10. ASI10: Rogue Agents
  11. Defense in Depth

Society and Open Problems

  1. Fairness
  2. AI Safety and Governance
  3. Open Problems in AI